Re: IIe faulty RAM ?

["Mark Cummings" <NOTfigjams@primus.com.au>]

Ed,

There is nothing haphazard about my trouble shooting techniques, I simply
pointed out all the things that had been done to proove that the motherboard
was faulty, because I knew I would be told to check things that I had
already.

BTW, I have been professionally trained in Logical Fault Finding Techniques
as part of an 18 month full time electronics technician course, as well as
numerous post graduate courses over the past 11 years.
The Logical Fault Finding Technique involves two steps which are performed
in sequence:
1. System Fault Finding
2. Component Fault Finding
System fault finding can be broken down into further steps: System,
Sub-System, Sub-Sub-System, & Stage, again performed in sequence.
For a IIe you would equate the System as the whole computer, incl disk drive
monitor, cpomputer box, power cord, etc. A Sub-System would be an individual
section like the computer box. A Sub-sub system would be an interface card,
a PSU, or a motherboard. And obviously a component would be one component on
a card.
The basic approach to Logical Fault Finding is to break the whole system
down into logical parts and use a half split technique to narrow down the
fault. In addition to that before beginning a another technique is used to
verify the that the procedure will work.
That is abbreveiated as C.E.P.I.O.
C=Calibration, meaning check the equipment you are going to use is
calibrated. In the case of a IIe, system fault finding there probably isn't
much electronic equipment required, perhaps a Multimeter or a Logic Probe.
Mine are not calibrated, but as I use it on other equipment regularly, I can
be reasonable confident that they at least work.
E=Eyeball, meaning have a quick look over the thing to see if anything
obvious is faulty, to save spending ages chasing your tail. Like the monitor
cord is not plugged into the right hole, or there is smoke or a buzzing
coming out of the power supply.
P=Power, all electronic equipment is designed to operate within specified
voltages. If you chase a fault and discover that it is caused by too high or
low voltage from the PSU, then the fault is in the PSU, not the equip.
I=Input, meaning appropriate inputs are applied to make it work. Is the
Power switched on at the wall. Is the power reaching the motherboard and
voltages correct, DC & AC.
O=Output, meaning you are getting the right output for a specified input.
This is the first step to give you the big picture. From there you can
decide where you will start to look for your fault. This is where guess work
or experience comes in, if you are confident you can get to the fault by
splitting the fault at a certain point, then do that, otherwise half-split.

I did use all the above steps to get to the point where I knew the
motherboard was faulty. Now I should go to the Component Fault Finding part.
With a computer, it is a little harder because of the way the busses are
shared, and signals are flying around everywhere simultaneously, however to
reduce the system to a minimum is advisable. Before that I checked for power
at the CPU and clocking pulses. They seemed OK. I then added a couple of
known good interfaces to get some more visual feedback as to how the was
affecting the whole system. The video card could have helped me diagnose a
faulty RAM condition, because occasionally the 80 column display came on and
I was looking for corruption on the alternate memory. In this case it didn't
help much, but I did have a logical reason for trying it. Adding the disk
drive was to see if the ROM was getting as far as booting. Without a card
fitted I wouldn't know. Putting the card in proved that it didn't.

At this stage I was thinking RAM, but it was soldered, so I opted to use a
simple elimination technique by swapping out the relativy few socketed
chips, in case I got lucky. I didn't. Trying to fault find the board to
component level would require a circuit diagram, which I don't have, except
for the one for a non-enhanced model. They are sufficiently different to
make it useless. So I use another techo's trick, by freezing and heating a
lot of IC's, especially RAM. In this case it doesn't reveal a fault, but
occasionally does. I am now down to a lot of parts swapping without a
diagram, so I decided to ask on the newsgroup who may have more experience
with specific faults on a IIe motherboard. Rather than bore everyone with
all of this above (like I am now, sorry) I put in the minimum facts required
and point out that the board is faulty.

So if you believe their is some error in my fault finding techniques, I
suggest you reveal your qualifications, and normal procedures, because I for
one, would be interested if there is a better technique. I have yet to see
one.

As far as the Kernal OK is concerned, that must have been for an un-enhanced
IIe or II+ as Igor point out, I can't check anymore, all mine are platinums.

Mark


<noone@nowhere.net> wrote in message 3B27841B.35DC2E8A@nowhere.net">news:3B27841B.35DC2E8A@nowhere.net...
> Mark:
> The infomraiton in your original post suggested haphazerd
> troubleshooting procedures.  I outlined a simple step by step procedure
> to narrow down the problem.  Nothing generic about that.
>
> The Kernel OK message is from a computer I have actually run it on.
> (Can't recall if that was my II+ or the Laser128,oh well.)
>
> Good luck with the ram replacement.
> Thankx,
> Ed
>
> Mark Cummings wrote:
> > thanks Ed, no offence, but I put a stack of info in my first post to cut
> > down the generic answers.